Malicious model weights and pickle-based exploits are emerging as serious threats to the AI supply chain. Here's how attackers weaponize model files, why safetensors matters, and what practitioners can do to verify the integrity of downloaded models.
The EU AI Act introduces tiered obligations that ripple across the AI stack, from foundation model providers to infrastructure firms. Here's what changes for compliance, transparency, and the synthetic media disclosure rules that matter most.
New research tackles CLIP's tendency to latch onto shortcut features, proposing a simplicity-bias-aware adaptation that improves generalizable deepfake detection across unseen generative models.
Latent space is the mathematical foundation behind generative models that power deepfakes, face swaps, and AI video. Here's how these compressed representations enable realistic synthetic media generation and manipulation.
A deep look at Higgsfield AI's role in the AI-generated film 'Cully Hill' reveals that despite the synthetic video pipeline, the most compelling elements still came from human writing, direction, and creative labor.
New projections warn that deepfake-enabled fraud and identity attacks could rise nearly 500 percent in 2026, driven by cheap, accessible generative tools and weak enterprise defenses.
Cryptographic attestation is emerging as a foundational layer for AI trust, letting systems prove where models run, what code executed, and whether outputs are authentic—with major implications for synthetic media provenance.
A clever new font remaps characters so human readers see normal text while AI scrapers ingest scrambled nonsense — a novel content-poisoning defense against unauthorized AI training and web crawling.
Guitar string maker D'Addario acknowledged using AI-generated music from Suno in a promotional video, triggering backlash from musicians and reigniting debate over synthetic audio in brand marketing.
Google DeepMind released SignGemma-powered tools that translate sign language into text in real time using on-device pose estimation and video recognition, marking a milestone for accessible AI-driven gesture understanding.
DETECT-World is billed as the first world model designed for deepfake detection, applying physics-aware scene understanding to spot inconsistencies that traditional pixel-level classifiers miss.
A new arXiv study moves beyond static detection to evaluate how defensive LLMs hold up against AI-generated social engineering in live, turn-by-turn conversations—exposing gaps in real-time defenses against synthetic manipulation.